Analyze Your Extracted Content

An assistant is the most visible way to use a dataset, but it isn't the only one. The content a data capture job captures is structured data you can pull out and analyze in your own tools. There are three ways to work with it, from most conversational to most analytical:

1. Answer With an Assistant

The primary path: connect the dataset to an assistant and let people ask questions in natural language. Best for day-to-day use, and for your visitors — it needs no technical setup on their side.

See the detail: Create an Assistant.

2. Export a Snapshot

From the dataset, export its contents as Excel, CSV, or JSON. This is a point-in-time file you can open in a spreadsheet to audit what was captured, hand to another system, or keep as a backup. Best for one-off analysis and archiving.

See the detail: Export a Dataset.

3. Connect a Live Feed (OData)

A dataset can also expose its contents as an OData feed — a queryable table that stays current. Point Excel (Power Query) or Power BI at the OData URL, load the ContentDetails table, and refreshing pulls the latest content without re-exporting anything. Best for dashboards and recurring analysis.

Available on the Enterprise plan. (Exporting a snapshot, above, is available on every plan.)

See the detail: Connect Excel with OData.

For example, a content team connects Power BI to a data capture job's dataset and builds a dashboard of every captured page — counts by section, last-updated dates, gaps in coverage. It refreshes on its own, so the dashboard always reflects the latest crawl instead of a stale export.